I. Massive and fast inversions provide high resolution inversion of MobileMT data into resistivity-depth models without limitations on frequency or station numbers. Results are used to assess data quality and compile databases for further analysis. II. Detail and goal-oriented inversions use adaptive finite element modeling focused on specific zones, providing more detailed models of complex structures. III. Forward modeling estimates the detectability of exploration targets using MobileMT, helping evaluate the technology's capability for specific geology.

1. Inversions
of MobileMT data
and forward
modelling
from innovations to discoveries
info@expertgeophysics.com www.expertgeophysics.com
I. Massive and fast inversions
II. Detail and goal-oriented
inversions
III. Forward Modeling

2. EM data inversion capability is the top of
EGL service offering
Except standard databases, grids and maps of acquainted data we
provide with inversion of EM data into resistivity-depth products –
sections, depth slices, 3d voxels:
2 MobileMT data inversions

3. 3 Massive and Fast Inversion
The first stage of the data inversion process is applying a nonlinear least-squares iterative
1d inversion algorithm based on the conjugate gradient method with the adaptive
regularization (M.Zhdanov, 2002). The algorithm uses weighting of the inverted
parameters, so sensitivity of the data to resistivity of each layer is remaining equal for
different depths. This way provides high resolution in the deep part of a model as well as in
the upper part. (the software development by N.Golubev)
An inversion is an ill-posed problem, and the algorithm uses regularization to get stable
and geologically meaning solution.We minimize parametric functional P which consists of
I. Massive and fast inversion
data misfit φ and stabilizer S (L2 norm of difference initial and fitted model) multiplied on parameter of
regularization α.
P(m)=φ(m)+αS(m) where
Φ(m)=‖dobs-dmod‖² is a data misfit and S(m)=‖m-mini‖² - stabilizer
The inversion is applied to each station without any resampling manipulations and without any limitations in
frequencies number and stations number along a line.
The fast inversion results are used for:
- Estimation of data quality and noise, if any, in particular frequency windows and stations;
- Compiling 3d resistivity-depth databases, voxels, 2d sections along lines, depth slices;
- Choosing anomaly zones and structures for detail and goal-oriented inversions;
- Developing optimal mesh-grids and starting model parametrization for further detail and goal-oriented
inversions.

4. 4 Massive and fast inversions. Examples
Examples

5. 5 Detail goal-oriented inversions footer
II. Detail and goal-oriented inversions
The next stage is detail and goal-oriented inversions based on adaptive finite elements and regularized
non-linear MARE2DEM, focusing on specific zones of interest (Kerry Key, Jeffrey Ovall. A parallel
goal‐oriented adaptive finite element method for 2.5‐D electromagnetic modelling. Geophysical Journal
International.Vol.186, Issue 1. 2011.).
The spatial domain fields are computed by solving the finite element system for a spectrum of
wavenumbers (about 30 logarithmically spaced wavenumbers), then using the inverse Fourier transform.
The used by MARE2DEM unstructured grids are very efficient for representing complex structures and
discrete targets.
MobileMT inverted section with unstructured grids

6. 6 Detail, goal-oriented Inversions. Examples
Examples
auto adaptively
and selectively
refined elements

7. 7 2/27/2020Add a footer
III. Forward Modeling
Objective of the forward modeling based on a geologic model and petrophysical parameters of the model is
estimation of detectability of an exploration target using MobileMT airborne EM.
Method: using MARE2DEM forward code to obtain MobileMT synthetic forward apparent
conductivity/resistivity data and then invert the calculated data with added noise into a resistivity section.
Forward modeling is a useful tool for feasibility of a method and equipment in geology characterization and in
a specific exploration task solving. EGL is open to investigate MobileMT capability with forward modeling for
qualitative estimation of the technology if clients provide with, even basic, geoelectric scenarios.
Below there is an example of forward modeling exercise for very challenging case for airborne EM systems
based on other principles.
Example
The model is taken from “Electromagnetic modeling based on the rock physics description of the true
complexity of rocks: applications to study of the IP effect in porphyry copper deposits” AbrahamM. Emond,
Michael S. Zhdanov, and Erich U. Petersen, University ofUtah. SEG/New Orleans 2006Annual Meeting.
This model scenario, developed by the authors, is typical in the southwestern U. S. and incorporates the
classic zones of a porphyry copper system. One of challenges for AEM is the presence of thick (>100 m)
conductive (6.5 S of conductance) overburden. The forward modeling results demonstrates great potential of
MobileMT in detecting these types of targets with the complicating conductive overburden factor.

8. 8 2/27/2020 Add a footer
Model and MobileMT apparent conductivity calculated response
mS/m

9. 9 2/27/2020 Add a footer
Inverted resistivity section with projected model boundaries
(unconstrained inversion with half space initial model)

10. Expert Geophysics Limited
info@expertgeophysics.com www.expertgeophysics.com